The rule is simple to state: when a plug enters a socket, the protective earth contact must be made before the live contacts, and broken after them. It is the reason the earth pin on a schuko or a British plug is longer, and the reason a two-pin plug with an earth clip is treated with suspicion.
What is less simple is that the rule is a comparison between two distances, each of which carries its own tolerance. A design that satisfies the rule at nominal dimensions can violate it at the extremes.
The two distances that matter
Distance one is how far the earth pin protrudes beyond the live pins. Distance two is how far the earth contact sits inside the socket relative to the live contacts. The first-make condition is satisfied when the earth travel to contact is shorter than the live travel, by enough margin to cover both tolerances.
| Item | Typical nominal | Tolerance that can eat the margin |
| Earth pin protrusion beyond live pins | 3 to 5 mm | ±0.3 mm on pin length |
| Earth contact depth in socket | 1 to 2 mm deeper | ±0.4 mm on contact position |
| Pin diameter (16 A) | 4.8 mm | −0.05 mm |
| Insertion force at first contact | 5 to 15 N | Spring set after 100 cycles |
| Worst-case margin | 0.3 mm or less | Inverted by tolerance stack |
How to measure it without a special rig
You do not need a high-speed recorder. A continuity tester on the earth circuit, a second one on the live circuit, and a slow manual insertion at about 20 mm/s will show the order: the earth lamp lights first, then the live lamp. Doing it five times on the same sample catches most problems.
For production, a simple Go fixture that blocks the plug until the earth has made contact is faster. The plug cannot be pushed home unless the earth circuit is closed first, so a plug with an inverted sequence physically cannot pass.
Where it goes wrong in practice
The common failure is not a bad design but a drifting process: the pin header that holds the three pins during moulding wears, one cavity sits 0.4 mm short, and the earth pin length distribution slowly shifts. The drawing is still right; the parts are no longer the drawing.
The second failure is assembly of the earth contact clip: a clip seated 0.5 mm high still makes contact, still passes a continuity test, and no longer guarantees first make. Measured as a position, not as continuity, it is obvious.
Standards referenced: IEC 60884-1 clause 9 (provision for earthing), IEC 60884-1 Annex A for the gauge dimensions used to check pin positions.
Why two-pin plugs with an earth clip are a separate case
A two-pin plug with a side earth clip still has to make earth first, and the geometry is worse: the clip is a flat spring that has to find a contact on the socket skirt, and its travel depends on how squarely the plug enters.
The usual arrangement gives the clip 2 to 3 mm of lead over the live pins. Tilt the plug 5 degrees on insertion and one side of the clip engages 0.3 mm earlier while the other engages 0.3 mm later, which can be enough to make one live pin touch first.
The measurement that shows the order
Two continuity testers, one on each circuit, and a slow insertion. What you are looking for is not the final state but the transition: earth on, then live on. Record the insertion distance at each transition with a rule against the plug body.
Repeat with the plug tilted 5 degrees in each direction. A design with real margin keeps the earth-first order in all four orientations; one with nominal-only margin loses it at the first tilt.
Where the requirement comes from
The purpose is arc containment. If the live contacts make before the earth, any arcing at the moment of contact happens in a circuit with no earth path, and the fault current has to find another route. With earth made first, the protective device operates on a defined path.
For Class I equipment this is a safety requirement, not a preference. For Class II it does not apply, which is why two-pin Class II products have no first-make requirement at all.
